Nonparametric Finite Time LTI System Identification

We address the problem of learning the parameters of a stable linear time invariant (LTI) system or linear dynamical system (LDS) with unknown latent space dimension, or order, from a single time--series of noisy input-output data. We focus on learning the best lower order approximation allowed by finite data. Motivated by subspace algorithms in systems theory, where the doubly infinite system Hankel matrix captures both order and good lower order approximations, we construct a Hankel-like matrix from noisy finite data using ordinary least squares. This circumvents the non-convexities that arise in system identification, and allows accurate estimation of the underlying LTI system. Our results rely on careful analysis of self-normalized martingale difference terms that helps bound identification error up to logarithmic factors of the lower bound. We provide a data-dependent scheme for order selection and find an accurate realization of system parameters, corresponding to that order, by an approach that is closely related to the Ho-Kalman subspace algorithm. We demonstrate that the proposed model order selection procedure is not overly conservative, i.e., for the given data length it is not possible to estimate higher order models or find higher order approximations with reasonable accuracy.

[1]  Benjamin Recht,et al.  Non-Asymptotic Analysis of Robust Control from Coarse-Grained Identification , 2017, ArXiv.

[2]  P. Wedin Perturbation bounds in connection with singular value decomposition , 1972 .

[3]  Yuanzhi Li,et al.  Even Faster SVD Decomposition Yet Without Agonizing Pain , 2016, NIPS.

[4]  Yi Zhang,et al.  Spectral Filtering for General Linear Dynamical Systems , 2018, NeurIPS.

[5]  Stephen Tu,et al.  On the Approximation of Toeplitz Operators for Nonparametric $\mathcal{H}_{\infty}$-norm Estimation , 2017, American Control Conference.

[6]  Benjamin Recht,et al.  Minimax Lower Bounds for $\mathcal{H}_\infty$-Norm Estimation , 2018 .

[7]  Samet Oymak,et al.  Non-asymptotic Identification of LTI Systems from a Single Trajectory , 2018, 2019 American Control Conference (ACC).

[8]  M. Meckes On the spectral norm of a random Toeplitz matrix , 2007, math/0703134.

[9]  Dietmar Bauer,et al.  Order estimation for subspace methods , 2001, Autom..

[10]  Munther A. Dahleh,et al.  On system identification of complex systems from finite data , 2001, IEEE Trans. Autom. Control..

[11]  Gábor Lugosi,et al.  Concentration Inequalities - A Nonasymptotic Theory of Independence , 2013, Concentration Inequalities.

[12]  Michael I. Jordan,et al.  Learning Without Mixing: Towards A Sharp Analysis of Linear System Identification , 2018, COLT.

[13]  Alexander Goldenshluger,et al.  Nonparametric Estimation of Transfer Functions: Rates of Convergence and Adaptation , 1998, IEEE Trans. Inf. Theory.

[14]  Lennart Ljung,et al.  System Identification: Theory for the User , 1987 .

[15]  Devavrat Shah,et al.  Model Agnostic Time Series Analysis via Matrix Estimation , 2018, Proc. ACM Meas. Anal. Comput. Syst..

[16]  Paul Tseng,et al.  Hankel Matrix Rank Minimization with Applications to System Identification and Realization , 2013, SIAM J. Matrix Anal. Appl..

[17]  Bart De Moor,et al.  Subspace Identification for Linear Systems: Theory ― Implementation ― Applications , 2011 .

[18]  Behçet Açikmese,et al.  Lossless Convexification of Nonconvex Control Bound and Pointing Constraints of the Soft Landing Optimal Control Problem , 2013, IEEE Transactions on Control Systems Technology.

[19]  Devavrat Shah,et al.  Time Series Analysis via Matrix Estimation , 2018, ArXiv.

[20]  Erik Weyer,et al.  Finite sample properties of system identification methods , 2002, IEEE Trans. Autom. Control..

[21]  E. Tyrtyshnikov A brief introduction to numerical analysis , 1997 .

[22]  Alexander Rakhlin,et al.  How fast can linear dynamical systems be learned? , 2018, ArXiv.

[23]  Roman Vershynin,et al.  Introduction to the non-asymptotic analysis of random matrices , 2010, Compressed Sensing.

[24]  D. Rubin,et al.  Maximum likelihood from incomplete data via the EM - algorithm plus discussions on the paper , 1977 .

[25]  Parikshit Shah,et al.  Linear system identification via atomic norm regularization , 2012, 2012 IEEE 51st IEEE Conference on Decision and Control (CDC).

[26]  Ambuj Tewari,et al.  Finite Time Identification in Unstable Linear Systems , 2017, Autom..

[27]  S. Geer,et al.  The Bernstein–Orlicz norm and deviation inequalities , 2011, 1111.2450.

[28]  S. Kung,et al.  Optimal Hankel-norm model reductions: Multivariable systems , 1980 .

[29]  T. Lai,et al.  Self-Normalized Processes: Limit Theory and Statistical Applications , 2001 .

[30]  Tengyu Ma,et al.  Gradient Descent Learns Linear Dynamical Systems , 2016, J. Mach. Learn. Res..

[31]  Anders Rantzer,et al.  Low-Rank Optimization With Convex Constraints , 2016, IEEE Transactions on Automatic Control.

[32]  Csaba Szepesvári,et al.  Improved Algorithms for Linear Stochastic Bandits , 2011, NIPS.

[33]  K. Glover All optimal Hankel-norm approximations of linear multivariable systems and their L, ∞ -error bounds† , 1984 .